Aug 26, 2009
Self-perfection by liquefaction shrinks waveguide loss
The light loss in an integrated optical pathway is often 20 times worse than in a conventional optical fibre due to edge roughness. As the width of the on-chip waveguide becomes smaller, the problem is exacerbated because more of the light is at the edge of the structure, which could be a major stumbling block for miniature optical circuits.
Considerable effort has been devoted to reducing the sidewall roughness to improve the performance of optical components. Techniques include anisotropic wet etching, thermal oxidation and combinations of the two. However, these methods are either limited to certain crystalline facets of semiconductor materials, or often involve harsh processing conditions, high temperatures or complicated procedures, which could degrade the substrate and other existing devices on the chip.
To solve these issues, researchers at Princeton University, US, have come up with a way of selectively smoothing the sidewall roughness of waveguides without stress or damage to either the existing devices or the substrate. Even better, the method is a self-perfection process and is simple to implement.
Self-perfection by liquefaction
Self-perfection by liquefaction (SPEL) is performed using a XeCl excimer laser with a wavelength of 308 nm and a pulse duration of 20 ns. The laser spot is about 3 mm × 3 mm and the laser pulse energy can be changed by adjusting a variable attenuator to selectively melt the surface layer of the waveguide. In a molten state, materials such as silicon have an extremely low viscosity (lower than the water), and can flow and smooth out the edge roughness by themselves under surface tension.
Using the technique, the Princeton researchers have reduced the average sidewall roughness of a silicon waveguide from 13 to 3 nm (1σ). According to their calculations, this is equivalent to reducing light loss from 53 to 3 dB/cm – a five orders of magnitude increase in light transmission for a 1 cm long waveguide.
Tests show that the new approach can fix defects within just 200 ns. This short exposure time together with SPEL's material selectivity, allows developers to repair defective components on a chip without damaging surrounding components and materials, and makes SPEL a promising candidate for enabling higher-density nanophotonics.
The researchers presented their work in Nanotechnology.
About the author
Dr Qiangfei Xia received his PhD in electrical engineering from Princeton University and is currently a research associate at Hewlett-Packard Labs in Palo Alto, CA. His research interests include nanoimprint lithography, nanofabrication and their applications in nanoscale devices and system integration. Dr Patrick F Murphy received his PhD in electrical engineering from Princeton University and is currently an associate at Exponent scientific and engineering consulting in New York, NY. Dr He Gao is a postdoctoral researcher at Carnegie Mellon University. He received his PhD from Princeton University and his research interests include novel nanofabrication processes and equipments, nanoscale devices for electronics, optics, data storage and renewable energy. Dr Stephen Y Chou is the Joseph C Elgin professor of engineering and head of the NanoStructure Laboratory at Princeton University. Dr Chou is a world leader, pioneer and inventor in a broad range of nanotechnologies, among which are nanoimprint lithography, quantized magnetic disks (bit-patterned media), ultra-small transistors and sub-wavelength optical elements.